Questions about scale and UVs

[JScullin]

Apologies if these answers are documented somewhere as I didn't see anything jump out on these particular issues.

Do the materials account for scene units when applied?
It seems they are metric but we're using U.S standard and notice the material is slightly larger and have to use an uvw xform set to .4 to .5 to match the scale scene in the vsphere scene. Is there a better workaround for this (without joining the rest of the world)?

UV's at real-world scale workflow?
Our workflow has our models unwrapped to RWS or sometimes UVW box mapped at RWS. So long as these are done correctly checking off "use real-world scale" will give us accurate placement? A lot of our processes are automated and unless we can script the auto-adjust feature our approach would be to slap these on and go.

Understanding mapping issues errors
Can I get more explanation on what is needed to correct "The U and V directions are not orthogonal" message? The angle seems to vary depending on object.
Also, when doing a detect issues a "bad mapping detected" with a percentage is given for objects in scene. Not sure how to interpret that.

Thanks in advance.

[Ivaylo Katev]

When the material is applied, it just obeys the uv mapping, ignoring the physical size. The connection with the real world is possible using the UV adjustment tool or using triplanar mapping.
The scanned materials do not prefer any unit system, in both the ways (the uv adjustment tool and the triplanar) the calculations are based on the scene unit, and you can set it to whatever you need. I have no idea why you need to apply 0.4 multiplier to achieve correct sizes, have to investigate it, but more details will be needed (the material, some scene, etc)
The message "The U and V directions are not orthogonal" appears when the UV mapping of the picked point is skewed with angle bigger than 2 degrees. Perhaps you miss one important detail here, when the tool asks you to pick the geometry, it is not just to tell which geometry is using this material, in fact both Maya and Max provide this info, and there is no need to ask the user. The goal is to pick a point that is considered "well mapped". For example, if you create simple sphere with the default mapping in Max, the points near the "equator" are well mapped, and the message will not appear, but the poles are "singularities" in the UV space, and if you pick a point near the poles the message will appear. In any case the presence of the message is disturbing - even if you pick a point with good mapping, the presence of points with bad mapping may cause artifacts with some anisotropic materials. Keep in mind that all the materials that are direction sensitive require better mapping quality, because they are using not only the UV coordinates of the hit point, but the U and V axes as well.
About the bad mapping percentage - when the UV checker is applied, every triangle in the mesh is checked for UV skew and if the value is above the specified angle, the surface of the triangle is added to the "bad surface" counter. After the entire mesh is checked, the bad surface is reported as percentage of the entire mesh surface. In fact, every curved surface has some UV skew, this is mathematically inevitable. So don't be scarry when all the geometry is reported with some percentage of bad mapping, this tool is used to attract the attention when the values are unusual.

[JScullin]

Thanks for that clarification! Yes, it seems our internal shaderballs are scaled wrong for vrscans and that may be the issue. There may have been a slight rescaling issue w/moving the vrsphere from the original scene to ours as well as it's working properly now.

Part of me was under the impression vrscans were apply & go and took the UV process out of the equation.

[Ivaylo Katev]

Part of me was under the impression vrscans were apply & go and took the UV process out of the equation.

This was considered to be the default behavior, but the silent actions may lead to problems.